This project is designed to determine the reliability and validity of questionnaire components in relation to physiologic measures of physical activity and fitness. This study consists of (1) a retrospective analysis of physical activity data collected biannually in the BLSA to describe change in physical activity levels over time and variation by age and sex, and the questions most strongly associated with treadmill-assessed fitness; (2) a prospectively-administered questionnaire to assess reliability of retrospective recall of physical activity as compared with previous self-report; and (3) prospective discrimination of the relationships between treadmill-measured fitness and daily activity (as measured by activity monitor and activity diary) with questionnaire assessment of habitual physical activity. This study will result in development of a questionnaire to assess physical activity levels in the BLSA population that has optimal validity for adults of all ages, and for both men and women. Development of such an instrument will maximize the validity of studies on the relationship between habitual physical activity levels and aging, health, and disease.